A novel histone exchange factor, protein γ, mediates the exchange …repository.cshl.edu/22839/1/A...

TH

EJ

OU

RN

AL

OF

CE

LL

BIO

LO

GY

JCB: ARTICLE

© The Rockefeller University Press $8.00The Journal of Cell Biology, Vol. 175, No. 3, November 6, 2006 389–400http://www.jcb.org/cgi/doi/10.1083/jcb.200608001

JCB 389

IntroductionIn eukaryotic nuclei, DNA is wrapped around a protein octamer

containing two copies of each of the core histones H2A, H2B,

H3, and H4, forming nucleosomes, the fundamental units of

chromatin (Luger et al., 1997). Nucleosomes are assembled with

the assistance of chaperones or assembly complexes. During

de novo nucleosome assembly, DNA is fi rst wrapped around the

H3–H4 tetramer before the addition of two H2A–H2B dimers

(Loyola and Almouzni, 2004). Once assembled, these core

histones are tightly bound to DNA, and the interactions must be

loosened or remodeled to allow the access of molecular

machineries (e.g., polymerases) to DNA. In living cells, the

histone–DNA interaction and chromatin structure are expected

to be continually altered during transcription, genome duplica-

tion, and damage recovery, and the remodeling of chromatin is

often associated with histone exchange (Belotserkovskaya and

Reinberg, 2004; Flaus and Owen-Hughes, 2004; Loyola and

Almouzni, 2004; Henikoff and Ahmad, 2005; Kimura, 2005).

On the other hand, nucleosome contexts on specifi c loci must be

preserved to maintain epigenetic marks on histone tails (Turner,

2002). The modifi cation of histones, including acetylation,

methylation, and phosphorylation, plays essential roles in chro-

matin functions such as gene expression and chromosome seg-

regation. Thus, fl uidity and stability seem to be well balanced in

nucleosomes in living cells.

Early studies of histone deposition and exchange in living

cells used radiochemical labeling. The stable association of

[3H]arginine-labeled H3–H4 with chromatin was demonstrated

by cell fusion (Manser et al., 1980). In a series of studies ana-

lyzing the deposition of the newly synthesized radio-labeled

histones into nucleosomes, it was found that linker histone

A novel histone exchange factor, protein phosphatase 2Cγ, mediates the exchange and dephosphorylation of H2A–H2B

Hiroshi Kimura,1 Nanako Takizawa,1 Eric Allemand,3 Tetsuya Hori,4 Francisco J. Iborra,5 Naohito Nozaki,6

Michiko Muraki,1,7 Masatoshi Hagiwara,7 Adrian R. Krainer,3 Tatsuo Fukagawa,4 and Katsuya Okawa2

1Nuclear Function and Dynamics Unit and 2Biomolecular Characterization Unit, Horizontal Medical Research Organization, Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto 606-8501, Japan

3Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 117244Department of Molecular Genetics, National Institute of Genetics and the Graduate University for Advanced Studies, Mishima, Shizuoka 411-8540, Japan5Medical Research Council Molecular Haematology Unit, Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, Headington, Oxford OX3 9DS, England, UK6Kanagawa Dental College, Yokosuka, Kanagawa 238-8580, Japan7Graduate School of Biological Science and Medical Research Institute, Tokyo Medical and Dental University, Bunkyo-ku, Tokyo 113-8510, Japan

In eukaryotic nuclei, DNA is wrapped around a protein

octamer composed of the core histones H2A, H2B, H3,

and H4, forming nucleosomes as the fundamental units

of chromatin. The modifi cation and deposition of specifi c

histone variants play key roles in chromatin function. In

this study, we established an in vitro system based on

permeabilized cells that allows the assembly and ex-

change of histones in situ. H2A and H2B, each tagged

with green fl uorescent protein (GFP), are incorporated

into euchromatin by exchange independently of DNA

replication, and H3.1-GFP is assembled into replicated

chromatin, as found in living cells. By purifying the

cellular factors that assist in the incorporation of H2A–

H2B, we identifi ed protein phosphatase (PP) 2C γ sub-

type (PP2Cγ/PPM1G) as a histone chaperone that binds

to and dephosphorylates H2A–H2B. The disruption of

PP2Cγ in chicken DT40 cells increased the sensitivity

to caffeine, a reagent that disturbs DNA replication

and damage checkpoints, suggesting the involvement

of PP2Cγ-mediated histone dephosphorylation and ex-

change in damage response or checkpoint recovery in

higher eukaryotes.

Correspondence to Hiroshi Kimura: [email protected]

Abbreviations used in this paper: AUT, acid-urea-Triton; FACT, facilitating chromatin transcription; Nap, nucleosome assembly protein; PB, physiological buffer; PP, protein phosphatase.

The online version of this article contains supplemental material.

on Decem

ber 13, 2011jcb.rupress.org

Dow

nloaded from

Published October 30, 2006

http://jcb.rupress.org/content/suppl/2006/10/30/jcb.200608001.DC1.html Supplemental Material can be found at:

http://jcb.rupress.org/

JCB • VOLUME 175 • NUMBER 3 • 2006 390

H1 and H2A–H2B undergo exchanges independently of DNA rep-

lication and transcription (Louters and Chalkley, 1985; Jackson,

1990). Such different behaviors of different histone species are

also seen in living cells using GFP-tagged proteins (Kimura,

2005). The linker histone H1 is rapidly exchanged within a few

minutes, and core histones are more stably bound. Long-term

observation and cell fusion experiments further revealed that a

substantial fraction of H2B-GFP exchanges slowly in euchro-

matin, whereas most H3-GFP (which is the H3.1 variant and is

referred to below as H3.1-GFP) and H4-GFP remain bound to

chromatin. In addition to the slowly exchanging fraction of

H2B-GFP, which exchanges independently of DNA replication

and transcription, another rapidly exchanging fraction, probably

coupled to transcription, has been observed (Kimura and Cook,

2001), which is in agreement with the dimer eviction observed

during transcription (Kireeva et al., 2002; Belotserkovskaya

and Reinberg, 2004).

The exchange and assembly of histones are regulated in

a development- and differentiation-specifi c manner (Meshorer

et al., 2006) by chaperones or assembly factors that are distinct

for each histone variant. Histone H3 has several variants, which

are deposited into specialized chromatin loci mediated differen-

tially through the action of deposition complexes (Loyola and

Almouzni, 2004; Tagami et al., 2004; Henikoff and Ahmad,

2005; Thiriet and Hayes, 2005). The modifi cation pattern in the

conserved tail is also distinctive in each variant (Hake et al.,

2006). H3.3 has modifi cations associated with transcriptionally

active chromatin, which is consistent with its localization on

active genes; in contrast, replication-coupled H3.2 has mostly

silencing modifi cations. Although the modifi cation pattern of

each histone is established after its assembly into nucleosomes

(infl uenced by the surrounding chromatin state), some specifi c

modifi cations are associated with nucleosome-free deposition

forms. Such modifi cations are typically found in H4, whose

deposition form is diacetylated throughout eukaryotes, and some

acetylation is associated with newly synthesized H3 in human

cells (Benson et al., 2006).

In addition to H3, variant-specifi c deposition and modifi -

cation are found in H2A. Nucleosome assembly protein 1 (Nap1)

and the related proteins are known as somatic H2A–H2B

chaperones after Nap1’s purifi cation from HeLa cells on the

basis of nucleosome assembly activity in vitro (Ishimi et al., 1984;

Loyola and Almouzni, 2004; Henikoff and Ahmad, 2005).

Although Nap1 is not essential for yeast viability, its disruption

affects the expression level of �10% of genes in clusters, sug-

gesting a nucleosome maintenance function of Nap1 by depos-

iting H2A–H2B (Ohkuni et al., 2003). Although Nap1 assists

nucleosome assembly without ATP in vitro, complexes contain-

ing ATP-dependent chromatin remodeling activity have recently

been shown to mediate the exchange of H2A–H2B dimers

(Bruno et al., 2003; Kobor et al., 2004; Krogan et al., 2004;

Mizuguchi et al., 2004). A complex containing yeast SWR1

(Swi2/Snf2-related ATPase 1) exchanges canonical H2A with

H2AZ in nucleosome arrays, and SWR1 and H2AZ regulate an

overlapping subset of genes. Another complex containing Tip60

(Tat-interacting protein 60) is involved in the exchange of phos-

phorylated H2Av (a Drosophila melanogaster histone H2A

variant homologous to H2AX) with the unphosphorylated form

at DNA lesions in Drosophila (Kusch et al., 2004). Thus, multi-

ple mechanisms appear to exist to control the exchange of H2A

variants at appropriate chromatin loci and in response to various

stimuli, including DNA damages.

To understand the molecular mechanisms that regulate the

assembly and exchange of histones in higher eukaryotes, we set

Figure 1. Visualization of histone exchange and assembly by permeabi-lized cells incubated in cell extract containing GFP-tagged histones. (A) Strategy. HeLa cells are permeabilized and incubated in extract pre-pared from cells expressing histone-GFP. Cy3-dUTP is incorporated into replicated chromatin. (B) Localization of histone-GFP in permeabilized cells incubated in cell extracts. Permeabilized cells were incubated in S100 ex-tract prepared from cells expressing GFP-H2A or H3.1-GFP. After washing and fi xation, DNA was counterstained with DAPI. Four views of single con-focal sections are shown. Insets show magnifi ed views of the boxed areas. (C) Localization of GFP-H2A in euchromatin. After incubating permeabilized cells with GFP-H2A–containing extract and Cy5-dUTP, cells were immuno -l abeled with rabbit polyclonal antibody specifi c to hyperacetylated H4 (top) or K20-trimethylated H4 (bottom) and Cy3-conjugated donkey anti–rabbit IgG. Typical examples of cells outside the S phase (without Cy5-dUTP incorporation) are shown. (B and C) Bars, 10 μm. (D) GFP-H2A replaces H2A in chromatin. Permeabilized cells were incubated in S100 extract from HeLa cells or cells expressing GFP-H2A. Mononucleosomes (input) and GFP-containing nucleosomes precipitated using anti-GFP (IP) were separated by SDS-PAGE. The line intensity profi les of histones stained with Coomassie are shown.

on Decem


Dow

nloaded from



A NOVEL HISTONE H2A–H2B EXCHANGE FACTOR • KIMURA ET AL. 391

out to establish an in vitro system that mimics in vivo histone

dynamics using permeabilized cells. When cells are treated with

nonionic detergents such as Triton X-100 or saponin, cellular

membranes are permeabilized and some proteins are extracted,

but many nuclear functions remain active (Jackson and Cook,

1985; Pombo et al., 1999), and some nuclear structures can be

manipulated by adding exogenous factors (Misteli and Spector,

1996; Maison et al., 2002). Therefore, we expected that exoge-

nously added histones might be incorporated into chromatin in

permeabilized cells by exchange or replication-coupled assembly.

As expected, GFP-tagged histones were indeed incorporated

into chromatin in permeabilized cells with the assistance of cel-

lular factors. By purifying the factors assisting GFP-H2A–H2B

incorporation, we identifi ed the type 2C protein phosphatase

(PP) 2Cγ/PPM1G (Travis and Welsh, 1997; Murray et al., 1999)

in addition to the Nap1 family members. PP2Cγ directly bound

to and dephosphorylated H2A–H2B, and its disruption in chicken

DT40 cells caused hypersensitivity to checkpoint abrogation.

Although PP2Cγ did not appear to be the major phosphatase for

H2AX and H2B, dephosphorylation and exchange via PP2Cγ

may function to allow full recovery from DNA damage.

ResultsHistone assembly and exchange in permeabilized cellsAs illustrated in Fig. 1 A, HeLa cells were permeabilized and

incubated in cell extracts prepared from cells stably expressing

GFP-tagged core histones whose expression levels were <10%

of their endogenous counterparts (Kimura and Cook, 2001). After

washing out the unincorporated materials, GFP-H2A localized

to euchromatin, which was devoid of DAPI-dense heterochro-

matin (Fig. 1 B, inset), in most permeabilized cells (Fig. 1 B). In

contrast, H3.1-GFP highlighted replicated chromatin, which was

labeled with Cy3-dUTP. The euchromatic localization of GFP-

H2A was confi rmed by the overlapping signals with specifi c

antibodies recognizing acetylated histone H3 (not depicted) and

H4 (Fig. 1 C), which are associated with transcriptionally active

chromatin (Turner, 2002). In contrast, GFP-H2A was excluded

from inactive chromatin rich in K20-trimethylated H4 (Fig. 1 C;

Turner, 2002). To confi rm that GFP-H2A replaced the endo-

genous H2A in chromatin in permeabilized cells, GFP- containing

mononucleosomes were prepared by immunoprecipitation using

antibody directed against GFP, and the ratio of core histones was

analyzed by SDS-PAGE and Coomassie staining (Fig. 1 D). The

amount of H2A was roughly halved in GFP-H2A nucleosomes

(Fig. 1 D), suggesting the incorporation of a dimer of GFP-H2A

and H2B (GFP-H2A–H2B) into chromatin by exchange. This

stoichiometry is unlikely to be created by the nonspecifi c ag-

gregation of a GFP-H2A–containing histone octamer from the

extract onto chromatin, as H2A–H2B and H3–H4 are present in

different complexes in the chromatin-free fraction (see Fig. 5 B;

Chang et al., 1997; Tagami et al., 2004).

We next used specifi c inhibitors to examine whether the

incorporation of histones depends on transcription and/or DNA

replication in permeabilized cells. Most H2A–H2B appeared

to be exchanged independently of ongoing RNA polymerase II

transcription and DNA replication, as the incorporation of GFP-

H2A and H2B-GFP was still observed in the presence of

α- amanitin and aphidicolin, respectively (Fig. 2 and not

depicted). The incorporation of H3.1-GFP into chromatin was

coupled to DNA replication, as the signal almost disappeared in

the presence of aphidicolin (Fig. 2). These results are reminis-

cent of the different behaviors of H2A–H2B and H3.1–H4 ob-

served in living mammalian cells (Louters and Chalkley, 1985;

Jackson, 1990, Kimura and Cook, 2001; Benson et al., 2006).

Purifi cation of the activity required for histone H2A–H2B incorporation from HeLa cell extractTo analyze whether GFP-H2A–H2B dimer alone can be in-

corporated into chromatin, permeabilized cells were incubated

Figure 2. Effects of polymerase inhibitors on the incorporation of GFP histones. Permeabilized HeLa cells were incubated in extracts prepared from cells expressing GFP-H2A and H3.1-GFP as in Fig. 1. In some cases, inhibitors of RNA polymerase II (2 μg/ml α-amanitin) and/or DNA polymerase (5 μg/ml aphidicolin) were added. Three views of single con-focal sections are shown. Bar, 10 μm.

on Decem


Dow

nloaded from




with GFP-H2A–H2B purifi ed from HeLa cells expressing GFP-

H2A (Fig. 3). Although GFP-H2A–H2B alone failed to be in-

corporated, its incorporation was restored when supplemented

with HeLa cell extract (Fig. 3 C), suggesting the presence of

soluble factors required for H2A–H2B exchange in the extract.

By following the incorporation of GFP-H2A under a fl uores-

cent microscope, we purifi ed the activity required for H2A–H2B

exchange using column chromatography (Fig. 4 A). The purest

active fraction consisted of three major bands by SDS-PAGE

(Fig. 4 B). Mass spectrometry analysis identifi ed these polypep-

tides as PP2Cγ/PPM1G (Travis and Welsh, 1997; Murray et al.,

1999), Nap1/Nap1L1 (Ishimi et al., 1984), and Nap2/Nap1L4

(Rodriguez et al., 1997). It was not surprising to fi nd Nap1

and Nap2 in the active fractions, as they have been described

as histone chaperones that assist nucleosome assembly in vitro

(Ishimi et al., 1984; Rodriguez et al., 1997). In contrast, no link

between PP2Cγ and histones was previously established, which

prompted us to focus on the function of PP2Cγ in histone H2A–

H2B exchange. The phosphatase might regulate the chaperone

activity of Nap1/2 by altering their phosphorylation state. On

the other hand, PP2Cγ might also mediate the H2A–H2B ex-

change as such because it has a unique acidic domain (Travis

and Welsh, 1997) that could potentially interact with histone

H2A–H2B.

Recombinant Nap1, Nap2, and PP2C𝛄 individually support H2A–H2B incorporationTo examine the relationship between H2A–H2B exchange ac-

tivity and the individual proteins, we incubated permeabilized

cells with the purifi ed GFP-H2A–H2B and each recombinant

protein fused to a histidine hexamer (His) tag expressed in and

purifi ed from Escherichia coli (Fig. 4, C and D). GFP-H2A was

incorporated into chromatin in the presence of either His-Nap1,

-Nap2, or -PP2Cγ (Fig. 4 D), and similar results were obtained

when ATP was omitted from the system (Fig. S1, available at

Figure 3. The incorporation of purifi ed GFP-H2A–H2B into permeabilized cell chromatin requires HeLa cell extract. (A) Strategy. Permeabilized HeLa cells are incubated in the nucleotide mixture containing Cy3-dUTP with GFP-H2A–H2B dimer in the presence or absence of extract from nonfl uorescent HeLa cells. (B) SDS-PAGE. H2A–H2B and GFP-H2A–H2B dimers purifi ed from chromatin of HeLa cells and cells stably expressing GFP-H2A, respectively, were analyzed by SDS-PAGE and Coomassie staining. (C) GFP-H2A incorporation. Permeabilized cells were incubated with 10 μg/ml of purifi ed GFP-H2A–H2B ± HeLa S100 extract. Two views (GFP and DAPI) of single confocal sections are shown. Bar, 10 μm.

Figure 4. Identifi cation of histone exchange activity. (A) Chromatography procedure to purify the activity assisting H2A–H2B exchange. (B) Identifi ca-tion of proteins associated with H2A–H2B exchange activity. The Superose 6 fractions were separated by SDS-PAGE and stained with Coomassie (top), and the activity supporting GFP-H2A incorporation was followed under a microscope (bottom). The 76- and 54/50-kD bands were identi-fi ed as PP2Cγ and Nap1/Nap2 by mass spectrometry as indicated. (C) Recombinant proteins. His-tagged PP2Cγ, the mutants, Nap1, and Nap2 expressed in and purifi ed from E. coli were analyzed by SDS-PAGE and Coomassie staining. (D) GFP-H2A–H2B incorporation supported by Nap1, Nap2, and PP2Cγ. Permeabilized cells were incubated in 10 μg/ml GFP-H2A–H2B ± HeLa extract or indicated recombinant proteins (30 or 90 μg/ml). In the case of mixing three proteins, an equal amount of each protein (10 or 30 μg/ml) was added to give the fi nal concentration (30 or 90 μg/ml, respectively). Two views of single confocal sections are shown. (B and D) Bars, 10 μm.

on Decem


Dow

nloaded from




http://www.jcb.org/cgi/content/full/jcb.200608001/DC1), sug-

gesting that PP2Cγ itself possesses ATP-independent chaper-

one activity, as do Nap1 and Nap2. Although the cofractionation

of these three proteins by gel fi ltration chromatography (Fig. 4 B)

suggests their presence in a complex, only additive effects on

the incorporation of GFP-H2A–H2B were observed when these

three recombinant proteins were mixed (Fig. 4 D).

As the acidic domain is unique to PP2Cγ among PP2C

family members (Travis and Welsh, 1997; Murray et al., 1999),

this domain might be essential for the chaperone function.

Indeed, a phosphatase mutant lacking the acidic domain (∆AcDo)

did not support GFP-H2A incorporation (Fig. 4 D). Furthermore,

coimmunoprecipitation analysis confi rmed that the physical

interaction of PP2Cγ with H2A–H2B requires this domain

(Fig. 5 A). When FLAG-tagged phosphatase was transiently ex-

pressed in human 293T cells and recovered using anti-FLAG

agarose beads, substantial amounts of endogenous H2A and

H2B were coprecipitated with FLAG-PP2Cγ but not with

FLAG-∆AcDo (Fig. 5 A). The interaction between basic pro-

teins like histones and the acidic domain could occur through

nonspecifi c binding as a result of the positive and negative

charges. However, the immunoprecipitation experiments show

specifi c binding of the phosphatase to H2A–H2B because only

these two, but not the other histones (i.e., H1, H3, and H4),

were coprecipitated even though all histone subtypes are posi-

tively charged. The complex formation between GFP-H2A–

H2B and PP2Cγ in the cell extract (used in Fig.1) was observed

by immunoprecipitation using anti-GFP agarose beads (Fig. 5,

B and C). The presence of PP2Cγ as well as Nap1 in the immuno-

precipitates was confi rmed by mass spectrometry (Fig. 5 B) and

immunoblotting (Fig. 5 C). A two-hybrid cDNA library screen

also yielded histone H2B as an interactor with PP2Cγ, and the

interaction required the phosphatase’s acidic domain (unpub-

lished data).

PP2C𝛄 dephosphorylates nucleosome-free histone H2A–H2BThe aforementioned results showing the physical interaction

between PP2Cγ and H2A–H2B suggest that these histones

could be substrates for the phosphatase. Therefore, we analyzed

the phosphorylation state of FLAG-PP2Cγ–bound histones us-

ing acid-urea-Triton (AUT) gel electrophoresis and immuno-

blotting with specifi c antibodies directed against phosphorylated

histones (Fig. 6 A). As expected, histones coprecipitated with

the wild-type phosphatase were poorly recognized by antiphos-

phohistone antibodies. In contrast, histones bound to a phos-

phatase-inactive mutant (D496A) comprised detectable levels

of phosphorylated molecules, including those related to DNA

damage response and apoptosis such as Ser139-phosphorylated

H2AX (called γ-H2AX; Rogakou et al., 1999) and Ser14-phos-

phorylated H2B (Cheung et al., 2003; Fernandez-Capetillo

et al., 2004), although the overall migration pattern was similar to

those bound to the wild-type phosphatase. As bulk nucleosomal

histones were still phosphorylated in cells overexpressing the

wild-type phosphatase (Fig. 6 A and by immunofl uorescence;

not depicted), only nucleosome-free H2A–H2B may be de-

phosphorylated by the phosphatase. Consistently, the purifi ed

His-PP2Cγ effi ciently dephosphorylated nucleosome-free his-

tones, including γ-H2AX in vitro (Fig. 6, C and D). As the D496A

mutant still supported histone exchange in permeabilized cells

(Fig. 4 D), the histone exchange and dephosphorylation do not

appear to be coupled. These results suggest that a nucleosome-

free H2A–H2B that binds to PP2Cγ may be dephosphorylated

before its deposition into a nucleosome. Although we did not

obtain positive data indicating the dephosphorylation of nucleo-

somal histones by PP2Cγ in overexpression and in vitro assays,

it is also possible that additional cellular factors, which may be

limited in the assays, stimulate the phosphatase activity or tar-

geting toward the nucleosomal histones.

We next tested whether PP2Cγ has in vitro nucleosome

assembly activity using a supercoiling assay (Fig. S2, available at

http://www.jcb.org/cgi/content/full/jcb.200608001/DC1) in which

the assembly of nucleosomes can be assessed by the formation

of supercoils from relaxed circular DNA (Ishimi et al., 1984;

Figure 5. Binding of histones with PP2C𝛄. (A) SDS-PAGE analysis of anti-FLAG immunoprecipitation. Immunoprecipitates from 293T cells trans-fected with GFP (control), FLAG-PP2Cγ (wild type [wt]), and FLAG-PP2Cγ (∆AcDo) were separated by SDS-PAGE and stained with Coomassie. The two bands smaller than 19 kD that coprecipitated with FLAG-PP2Cγ were identifi ed as histone H2B and H2A by mass spectrometry and immunoblot-ting. (B and C) Immunoprecipitation of GFP-H2A and its binding proteins. GFP-H2A was immunoprecipitated from S100 extract using anti-GFP aga-rose beads. The immunoprecipitated materials from control (HeLa) and ex-tract prepared from cells stably expressing GFP-H2A (HeLa:GFP-H2A) were separated by SDS-PAGE and stained with Coomassie (B) or analyzed by immunoblotting (C). (B) Coomassie staining. The positions of size stan-dards are indicated on the left. Proteins identifi ed by mass spectrometry are indicated on the right. Asterisks indicate IgG heavy and light chains re-leased from beads. (C) Immunoblotting. The membranes were probed with anti-PP2Cγ (left) and anti-Nap1 (right). The target proteins of the antibodies and IgG from beads are indicated by arrows and asterisks, respectively.

on Decem


Dow

nloaded from




Rodriguez et al., 1997). Most of the plasmid DNA became su-

percoiled in the presence of Nap1, but only some supercoiled

molecules accumulated even in the presence of high levels of

PP2Cγ (Fig. S2), indicating that PP2Cγ has only weak de novo

nucleosome assembly activity.

Effect of PP2C𝛄 knockdown on H2A–H2B mobility in living HeLa cellsTo investigate whether PP2Cγ is involved in the regulation of

H2A–H2B kinetics in living cells, we knocked down PP2Cγ in

HeLa cells expressing histone-GFP using RNAi; 3 d after the

transfection of a specifi c siRNA, the level of PP2Cγ decreased

substantially to <5% of the normal level (Fig. 7, A and B).

The mobility of H2A–H2B was analyzed by fl uorescence recovery

after photobleaching (Kimura and Cook, 2001). The recovery

kinetics of both GFP-H2A and H2B-GFP decreased in cells

transfected with PP2Cγ-specifi c siRNA compared with those

with the control siRNA (Fig. 7, C and D), whereas the mobility

of the linker histone H1c-GFP was unaffected (Fig. 7 E). These

observations in living cells refl ect the results from in vitro

assays, suggesting that at least a part of H2A–H2B exchange

is mediated by PP2Cγ as a histone chaperone in HeLa cells.

PP2C𝛄-defi cient DT40 cells show hypersensitivity to caffeineTo gain further insights into the biological function of PP2Cγ in

vertebrate cells, we established PP2Cγ-defi cient chicken DT40

cells by gene targeting (Fig. S3, available at http://www.jcb.org/

cgi/content/full/jcb.200608001/DC1). As the defi cient cells

were generated by a simple knockout strategy to disrupt both

alleles, PP2Cγ does not appear to be essential for cell growth.

However, substantial growth defects were observed when DNA

replication and damage checkpoints were abrogated by caffeine,

which preferentially inhibits ataxia telangiectasia mutated– and

ataxia telangiectasia and RAD3 related–dependent pathways,

although its exact interfering points remain elusive (Kaufmann

et al., 2003; Abraham, 2004). As shown in Fig. 8, PP2Cγ-

defi cient cells were more sensitive to caffeine compared with

the wild type in a growth rate assay (Fig. 8 A) and in a colony

formation assay (Fig. 8 B). In 2 mM caffeine, the wild-type

cells continued to grow for 3 d, but PP2Cγ-defi cient cells

stopped growing at day 2. At a higher concentration (4 mM), the

number of live cells (judged by the exclusion of trypan blue)

became considerably lower after day 2 in PP2Cγ-defi cient cells

(Fig. 8 A). The colony formation assay revealed that the survival

rate after 22 h of incubation in 4 mM caffeine was 35 ± 8 and

8.2 ± 0.3% in the wild-type and PP2Cγ-defi cient cells, respec-

tively (Fig. 8 B). As caffeine is known to sensitize cells to DNA

double-strand breaks induced by ionizing radiation (Kaufmann

et al., 2003; Abraham, 2004), we compared the sensitivity of

these cells with γ-ray irradiation in the presence or absence of

caffeine. Although PP2Cγ-defi cient cells showed a similar radi-

ation sensitivity to the wild type without caffeine, they became

more sensitive when 1 mM caffeine was present in the colony-

forming medium (Fig. 8 C).

These results indicate that PP2Cγ is not essential for DNA

double-strand break repair but suggest its involvement in recov-

ery from damage. As H2AX is known to be phosphorylated

around damaged chromatin, its dephosphorylation is required

for full recovery from the damage response (Chowdhury et al.,

2005; Keogh et al., 2005). Even though PP2A is likely to be the

major γ-H2AX phosphatase in higher eukaryotes (Chowdhury

et al., 2005), PP2Cγ could be involved in a backup dephosphor-

ylation and deposition pathway. To assess the role of PP2Cγ in

γ-H2AX dephosphorylation, the phosphorylation level of

H2AX (i.e., the signal detected with antibody directed against

γ-H2AX) was compared between the wild-type and PP2Cγ-

defi cient cells in response to DNA damage combined with treat-

ment with calyculin A, an inhibitor of PP1 and PP2A (Nazarov

et al., 2003; Chowdhury et al., 2005). In both cells, γ-H2AX

Figure 6. Dephosphorylation of histones by PP2C𝛄. (A) PP2Cγ-bound his-tones analyzed by AUT gel electrophoresis. After transfecting FLAG-PP2Cγ (wild type [wt]) or FLAG-PP2Cγ (D496A) into 293T cells, PP2Cγ-bound and nucleosomal histones were separated in an AUT gel. Coomassie stain-ing (left) and immunoblots with the indicated phospho-specifi c antibodies are shown. The positions of histone subtypes and their phosphorylated forms (arrows) are indicated. (B) Dephosphorylation of histones by PP2Cγ. 2 μg of 32P-labeled histones, which were phosphorylated by MSK1, were incubated with different amounts (lanes 2 and 5, 12.5 ng; lanes 3 and 6, 50 ng; lanes 4 and 7, 250 ng) of His-PP2Cγ (wt; lanes 2–4) or mutant (D496A; lanes 5–7) and separated by SDS-PAGE. The radioactivity and Coomassie-stained histones are shown. The 32P signals from all of the his-tones disappear by dephosphorylation with increasing amounts of His-PP2Cγ but not with His-PP2Cγ (D496A). (C) Dephosphorylation of H2AX by PP2Cγ. 1 μg histone H2A–H2B fraction prepared from 12 Gy–irradi-ated HeLa cells was incubated alone (lane 1) or with different amounts (lanes 2 and 5, 1 ng; lanes 3 and 6, 5 ng; lanes 4 and 7, 25 ng) of His-PP2Cγ (wt; lanes 2–4) or mutant (D496A; lanes 5–7) and separated by SDS-PAGE. The phosphorylation level of H2AX was analyzed by immuno-blotting with anti–γ-H2AX. The Coomassie-stained gel (CBB) is shown as a loading control.

on Decem


Dow

nloaded from




appeared at a similar level 2 h after irradiation (8 Gy) and disap-

peared by 8 h (Fig. 8 D, lanes 1–6); faint signals of apoptosis-

associated H2B (S14) phosphorylation appeared by 8 h. When

cells were incubated with calyculin A, γ-H2AX was accumu-

lated by 8 h even in the wild-type cells, probably as a result of

spontaneous or replication-associated damages, which is con-

sistent with the involvement of PP2A in γ-H2AX dephosphory-

lation (Chowdhury et al., 2005). The levels of γ-H2AX and

phospho-H2B (S14) were higher in PP2Cγ-defi cient cells in the

presence of calyculin A (Fig. 8 D, lanes 9 and 12), suggesting

that PP2Cγ is also one of the phosphatases that regulate

γ-H2AX and phosphorylated H2B. The difference between wild-

type and mutant cells became more evident when the cells were

irradiated and incubated in calyculin A, as more γ-H2AX and

phospho-H2B (S14) were accumulated in PP2Cγ-defi cient cells

(Fig. 8 D, lanes 13–18). These results suggest that PP2Cγ de-

phosphorylates γ-H2AX and phosphorylated H2B in wild-type

DT40 cells.

DiscussionIdentifi cation of histone chaperones required for H2A–H2B incorporation into chromatin in permeabilized cellsTo understand the biological function and molecular mecha-

nisms of histone dynamics, we established a permeabilized cell-

based assay for histone assembly and exchange. GFP-H2A and

H2B-GFP were incorporated into euchromatin in permeabilized

cells. This is consistent with the exchange of H2A–H2B in liv-

ing cells, which can occur independently of DNA replication

Figure 7. Effects of PP2C𝛄 knockdown on kinetics of GFP-tagged H2A and H2B in living cells. (A and B) Knockdown of PP2Cγ by siRNA. The amount of PP2Cγ in HeLa cells transfected with control or PP2Cγ-specifi c siRNA was evaluated by immunoblotting (A) and immunofl uorescence (B; 3 d after transfection) with anti-PP2Cγ. Anti–α-tubulin was used as a loading control. (C–E) FRAP. 3 d after RNA transfection, the mobility of his-tone-GFP was analyzed by bleaching a half of a nucleus (H2A and H2B) or a 2-μm spot (H1c) after the fl uorescence recovery. Examples (GFP-H2A) and the recovery curves of GFP-H2A and H2B-GFP (D) or H1c-GFP (E) are shown. The means of the relative intensity in the bleached area are indi-cated with the SD (n ≥ 9). (B and C) Bars, 10 μm.

Figure 8. Phenotypes of PP2C𝛄-defi cient cells. (A) Growth curve. The wild-type (Cl18; left) and PP2Cγ-defi cient cells (clone KO30; right) were plated (105 cells/ml), grown in caffeine (0, 2, or 4 mM), and the number of cells excluding trypan blue was counted every 24 h until 72 h. n = 4. (B and C) Colony formation assay. The number of colonies 10–12 d after plating was expressed as the relative value to that in controls without treat-ments. n = 3. (A–C) The mean and SD (error bars) are shown. (B) Caffeine sensitivity. Cells were treated with caffeine for 4, 10, or 22 h, diluted, and plated. (C) Radiation sensitivity. Cells were plated in methylcellulose medium ± 1 mM caffeine, irradiated (0, 2, 4, or 8 Gy), and incubated. (D) Effect of calyculin A on the phosphorylation of H2AX and H2B. The wild-type (wt; lanes 1–3, 7–9, and 13–15) or PP2Cγ-defi cient (lanes 4–6, 10–12, and 16–18) cells were irradiated (8 Gy; lanes 1–6 and 13–18) or not irradiated (lanes 7–12), and calyculin A was added (lanes 7–18). Cells were collected either immediately (0 h), 2, or 4 h after irradiation, and the levels of γ-H2AX and phosphorylated H2B (S14) were analyzed by immunoblotting. The Coomassie-stained gel (CBB) is shown as a load-ing control. KO, knockout.

on Decem


Dow

nloaded from




and transcription (Jackson, 1990; Kimura and Cook, 2001),

preferentially in chromatin-containing acetylated H4 (Benson

et al., 2006). H3.1-GFP assembled into replicated chromatin but

contrasted to H2A–H2B, which is also reminiscent of the be-

havior in living cells (Kimura and Cook, 2001). By purifying

the activity that assists GFP-H2A–H2B incorporation into chro-

matin in permeabilized cells, we identifi ed three proteins—

Nap1, Nap2, and PP2Cγ—in the purest fraction. Finding these

Nap1-related proteins in our active fractions reassures us that

the permeabilized cell-based assay has physiological relevance.

The third protein we found was PP2Cγ, which harbors a unique

acidic domain (Travis and Welsh, 1997) and was purifi ed as a

factor that stimulates spliceosome assembly in vitro (Murray

et al., 1999).

Our analyses indicated that the phosphatase as such can

assist the incorporation of H2A–H2B into chromatin in permea-

bilized cells and that it binds to and dephosphorylates histone

H2A and H2B subtypes. Although the acidic domain of PP2Cγ

could potentially mediate nonspecifi c electrostatic binding to

basic proteins such as the histones, the fact that H2A–H2B was

exclusively coimmunoprecipitated among all of the histones

using FLAG-tagged phosphatase suggests that the interaction

between PP2Cγ and H2A–H2B is specifi c. These histone chap-

erones do not require ATP for assisting H2A–H2B incorpo-

ration into chromatin in permeabilized cells as well as for in

vitro nucleosome assembly with naked DNA. Because we fol-

lowed the most active fractions that support GFP-H2A incor-

poration globally in euchromatin, other H2A–H2B exchange

factors that are probably less abundant and act on more spe-

cifi c loci, including facilitating chromatin transcription (FACT;

Belotserkovskaya and Reinberg, 2004) and ATP-dependent

remodeling factors (Flaus and Owen-Hughes, 2004), were not

found in the fi nal preparation. Although Nap1/2 and PP2Cγ

may mediate global H2A–H2B exchange independently of tran-

scription and DNA replication, FACT may participate in tran-

scription-coupled exchange. Future studies may reveal whether

FACT supports H2A–H2B incorporation in a transcription-

dependent manner in permeabilized cells.

A recent study revealed that ATP-dependent chromatin

remodeling complexes can mediate histone exchange in addi-

tion to their remodeling function without the displacement of

histone octamers (Flaus and Owen-Hughes, 2004). Therefore,

it is also possible that the function of ATP-independent chaper-

ones like Nap1/2 and PP2Cγ is solely to escort H2A–H2B and

transfer the dimer to the ATP-dependent machineries, such as

the yeast SWR1 complex that catalyzes the exchange between

a canonical dimer and an H2AZ–H2B dimer (Mizuguchi

et al., 2004). However, several lines of evidence suggest that

the chaperones might also mediate H2A–H2B incorporation

by themselves in addition to their escorting function. First,

yeast Nap1 has the ability to exchange H2A–H2B in mono-

nucleosomes in vitro (Park et al., 2005). Second, additional

ATP is not required for H2A–H2B incorporation supported

by Nap1/2 and PP2Cγ in permeabilized cells. Third, a sub-

stantial H2B-GFP recovery was still observed in living cells

by FRAP even when the cellular ATP pool was depleted by

treatment with sodium azide (unpublished data). Thus, although

ATP-dependent factors might be required for the exchange

of a dimer containing H2AZ at specifi c loci or during gene

activation, ATP-independent chaperones may participate in the

basal level of exchange of the major H2A and other variants.

Alternatively, the major role of ATP-independent chaperones

may be to deposit an H2A–H2B dimer into an incomplete

nucleosome lacking a dimer, which can result from positive

torsional stress (Jackson et al., 1994) or through ATP-driven

eviction. This may account for the slow exchange rate of

H2A–H2B in living cells despite the presence of a large pool

of PP2Cγ (�106 molecules/HeLa cell) diffusing almost freely

in the nucleus (unpublished data).

Involvement of PP2C𝛄 in DNA damage responseTo understand the biological function of PP2Cγ at the cellu-

lar level, we used chicken DT40 cells to create knockout cells

by gene targeting. Although the defi cient cells are viable, they

show subtle growth retardation and a remarkable hypersensi-

tivity to caffeine, which abrogates DNA replication and dam-

age checkpoints. One possible mechanism to explain these

phenomena is that the chaperone function together with the

phosphatase activity plays a role in completing chromatin

formation after DNA repair and/or replication by depositing

dephosphorylated H2A–H2B molecules (Fig. 9). H2AX is

phosphorylated around damaged chromatin (Rogakou et al.,

1999), and its dephosphorylation is required for full recovery

from damage responses. Also, H2AX molecules outside the

damaged area are kept from undergoing phosphorylation for

several hours. Although PP2A seems to play a major role in

γ-H2AX dephosphorylation on chromatin (Chowdhury et al.,

2005), we showed that PP2Cγ likewise mediates γ-H2AX

and H2B dephosphorylation, as PP2Cγ-defi cient cells showed

a greater accumulation of γ-H2AX and phosphorylated H2B

(S14) compared with wild-type cells when PP1 and PP2A were

inhibited by calyculin A.

Although the eviction of γ-H2AX or phosphorylated H2B

may be mediated by other proteins such as the Drosophila

Tip60-containing complex (Kusch et al., 2004), PP2Cγ may

passively deposit dephosphorylated H2A–H2B or H2AX–H2B

Figure 9. A model for PP2C𝛄 function. PP2Cγ binds to nucleosome-free H2A–H2B (or H2AX–H2B) and removes phosphate groups (indicated by circled P) before the next deposition. Alternatively, the phosphatase activity and/or substrate specifi city might be controlled by binding with histones. See Discussion for details.

on Decem


Dow

nloaded from




to incomplete nucleosomes lacking one dimer. This view is con-

sistent with the observed uncoupling of the chaperone function

and phosphatase activity of PP2Cγ; histone dephosphorylation

can occur at any time after the binding of PP2Cγ until deposition

(Fig. 9). Most H2A–H2B that bound to PP2Cγ but away from

nucleosomes was indeed dephosphorylated. The lack of PP2Cγ

in the DT40 knockout cells may thus delay the recovery from

damage. When checkpoints are functional, such a subtle repair

defect would not be critical and might only cause a subtle

delay in cell growth. However, when checkpoints are abrogated,

more cells with damaged chromatin would enter into mitosis

for catastrophe.

An alternative possibility is that the substrate specifi city

or phosphatase activity of PP2Cγ is regulated by binding to

H2A–H2B (Fig. 9); the level of nucleosome-free H2A–H2B

could be altered by damage or replication fork arrest. The type

2C phosphatase family members are indeed involved in check-

point responses (Leroy et al., 2003; Lu et al., 2005), and the

γ subtype in particular might take part in inactivating check-

points by sensing the free H2A–H2B level in the nucleus.

Finally, a link between chromatin-remodeling factors and alter-

native pre-mRNA splicing was recently reported (Batsche et al.,

2006). Consistent with this observation, PP2Cγ was previously

identifi ed as a factor that stimulates pre-mRNA splicing in vitro

(Murray et al., 1999), raising the interesting possibility that

PP2Cγ coordinately regulates stress responses in mammalian

cells at the level of chromatin and RNA splicing.

Concluding remarksIt is now widely acknowledged that histone modifi cation is key

for the regulation of chromatin functions. Recent studies further

indicate that the deposition and exchange of appropriate histone

variants to specifi c chromosome loci are also important for gene

expression and genome integrity (Loyola and Almouzni, 2004;

Henikoff and Ahmad, 2005). A connection between the histone

modifi cation and deposition has been shown typically in the

case of histone H4; before replication-coupled assembly, the

newly synthesized molecules are diacetylated by HAT1 histone

acetylase in the H3.1–H4 deposition complex (Chang et al.,

1997; Tagami et al., 2004). Although diacetylation is not a pre-

requisite for assembly (Ma et al., 1998), this modifi cation con-

tributes to the recovery from replication block-mediated DNA

damage (Barman et al., 2006). Similarly, in the case of H2A–

H2B and H2AX–H2B, the deposition of unphosphorylated

forms mediated by PP2Cγ appears to play a role in DNA dam-

age responses. Thus, controlling the incorporation of appropri-

ately modifi ed histones seems to be important for maintaining

genome integrity. Future studies should reveal how individual

ATP-independent chaperones and ATP-dependent remodeling

complexes function in distinct exchange processes in different

chromatin contexts. Although differences in histone exchange

kinetics in vivo were shown decades ago (Manser et al., 1980;

Louters and Chalkley, 1985), the biological signifi cance of the

exchange and the underlying molecular mechanisms are just

emerging. The approach presented in this study may contribute

to bridging the gap between live cell observations and biochem-

ical analyses.

Materials and methodsHistone exchange and assembly in permeabilized cellsIn typical experiments, HeLa cells were plated in a 12-well plate containing 15-mm coverslips and were grown up to subconfl uence. Cells were chilled on ice, washed twice in ice-cold physiological buffer (PB; 100 mM CH3COOK, 30 mM KCl, 10 mM Na2HPO4, 1 mM DTT, 1 mM MgCl2, and 1 mM ATP; Jackson and Cook, 1985) containing 5% Ficoll (PBF; pH 7.4; 1 ml per well; Nacalai Tesque), permeabilized in PBF containing 0.1% Triton X-100 (1 ml; for 5 min on ice), and washed twice in 1 ml PBF on ice. Cells were incubated for 1 h at 30°C in a reaction mixture containing cell extract (40%) or purifi ed proteins supplemented with 100 μM each of NTP and dNTP (GE Healthcare), 0.4 μM Cy3-dUTP (PerkinElmer), and 800 μM MgCl2 in PBF. For incubation, a coverslip was overlaid (cell side down) on a 100-μl drop of the reaction mixture on Parafi lm covering a fl at aluminum block in a water bath at 30°C. After washing twice in 1 ml PBF for 5 min on ice in a 12-well plate, cells were fi xed in 4% PFA (Electron Microscopy Sciences) in 250 mM Hepes-NaOH, pH 7.4 (Wako), for 20 min at room temperature, washed three times in 1 ml PBS, and DNA was counterstained with DAPI (12.5 ng/ml in PBS; 1 ml for 15 min; Nacalai Tesque). After washing twice in 1 ml PBS, coverslips were mounted using Prolong Gold (Invitrogen). In some cases, ATP and the other nucleotides were omitted from PBF and the reaction mixture.

For immunolabeling (Fig. 1 C), permeabilized cells were incubated in the reaction mixture containing 40% GFP-H2A extract and 2 μM Cy5-dUTP instead of Cy3-dUTP for 30 min at 30°C. After fi xation, cells were treated with 1% Triton X-100 in PBS for 20 min, washed fi ve times in PBS, and incubated in blocking buffer (0.2% gelatin, 1% BSA, and 0.05% Tween 20 in PBS, pH 8.0) for 30 min and then with rabbit polyclonal anti-bodies directed against hyperacetylated H4 (1:1,000; Upstate Biotechnol-ogy) or H4-trimethylated K20 (1:500; Abcam) in the same buffer for 3 h. Cells were washed in PBS containing 0.05% Tween 20 (PBST) three times for 10 min, incubated in Cy3-conjugated donkey anti–mouse IgG (1:500; Jackson ImmunoResearch Laboratories) overnight at 4°C, and washed with PBST three times for 10 min before DAPI staining.

Fluorescence images were sequentially collected using a confocal microscope featuring 405-, 488-, 543-, and 633-nm laser lines with the optimized pinhole setting operated by the built-in software: either a micro-scope (LSM510 META; Carl Zeiss MicroImaging, Inc.) with a C-Apo 40× NA 1.2 objective lens (for Figs. 1 B and 7 B) or a microscope (FV-1000; Olympus) with a UPlanSApo 60× NA 1.35 lens (for Figs. 1 C and 2–4). Image fi les were converted to tiff format using the operating software, merged, linearly contrast stretched (with the same setting in each set of experiments) using Photoshop version 7.01 (Adobe), and imported into Canvas 8 (Deneva) for assembly.

For chromatin immunoprecipitation, cells were centrifuged at 1,300 g for 10 min at 4°C after each step for buffer replacement. After the incuba-tion and washing, nucleosomes were prepared, and GFP-containing nu-cleosomes were precipitated as described previously (Kanda et al., 1998; Kimura and Cook, 2001).

Preparation of cell extracts and protein purifi cationHeLa cells and derivatives expressing H2B-GFP (Kanda et al., 1998) and H3.1-GFP were grown as described previously (Kimura and Cook, 2001), and lines expressing GFP-H2A and H1c-GFP were established by transfect-ing the expression vectors (Misteli et al., 2000; Perche et al., 2000). Cell extracts were prepared based on the study by Dignam et al. (1983) with modifi cations. The S100 extract was prepared using a 1.5 cell-packed vol-ume of 10 mM CH3COOK, 3 mM KCl, 1 mM Na2HPO4, 1 mM MgCl2, 1 mM ATP, 1 mM DTT, 10 mM Hepes-KOH, pH 7.4, and Complete protease inhibitor cocktail (EDTA-free; Roche) and dialyzed against PB plus inhibi-tors (1.5 μg/ml leupeptin, 2.5 μg/ml aprotinin, and 1 μg/ml pepstatin A; Wako). The nuclear pellet was extracted using an equal volume of 20 mM Hepes-KOH, 0.6 M KCl, 0.2 mM EDTA, 25% glycerol, 1 mM DTT, 1.5 mM MgCl2, and protease inhibitor cocktail (Roche) to yield the nuclear extract, which was also dialyzed against PB. Histone H2A–H2B and H3–H4 were separately purifi ed from the nuclear pellet essentially according to Simon and Felsenfeld (1979), and the GFP-H2A–H2B fraction was separated from untagged H2A–H2B using gel fi ltration column chromatography (HiLoad Superdex 75; GE Healthcare).

To purify the activity assisting histone H2A–H2B incorporation in permeabilized cells, S100 extract was fi rst fractionated through a histone H2A–H2B column, which was prepared by coupling 3 mg of the purifi ed H2A–H2B to 1 ml N-hydroxysuccinimide ester–activated Sepharose (GE Healthcare) according to the manufacturer’s instructions. Approximately

on Decem


Dow

nloaded from




6 mg/ml HeLa S100 extract was mixed with 5 M NaCl to yield a fi nal salt concentration of 0.5 M before applying to the column (2 ml per run). After washing with fi ve column volumes of PB containing 0.5 M NaCl, bound proteins were eluted with a linear gradient of NaCl (0.5–2 M in PB; 20 column volumes). Each fraction was concentrated, and the buffer was substituted to PB using Ultrafree 0.5 (Millipore) before use in the permeabilized cell assay with 10–20 μg/ml of purifi ed GFP-H2A–H2B. The activity supporting the nuclear localization of GFP-H2A except in nucleoli was followed under a fl uorescence microscope (Axiovert2; Carl Zeiss MicroImaging, Inc.). The active fractions eluted in �1 M NaCl were further separated using a MonoQ 5/50 GL column (GE Healthcare) with a linear gradient of NaCl (0–2 M in PB; 20 column volumes). The most active fraction eluted in �0.6 M NaCl was concentrated and sepa-rated (0.2 ml in each fraction) on a Superose 6 gel fi ltration column (GE Healthcare). Proteins were identifi ed by mass spectrometric analysis using a mass spectrometer (Ultrafl ex TOF/TOF; Bruker Daltonics) and by comparison between the determined molecular weights and theoretical peptide masses from the proteins registered in the National Center for Biotechnology Information.

Recombinant proteins and phosphatase assayThe cDNAs encoding human Nap1 (Nap1L1) and Nap2 (Nap1L4) were amplifi ed by reverse transcription (Revertra Ace and oligo-dT18; TOYOBO) of HeLa RNA and PCR (high fi delity PCR master; Roche) using the following primers designed from GenBank/EMBL/DDBJ acces-sion no. BT007023 (Nap1) and BC022090 (Nap2): Nap1 forward, T T A C C A T A T G G C A G A C A T T G A C A A C A A A G A A C A G T C ; Nap1 reverse, C A T C A A G C T T C A C T G C T G C T T G C A C T C T G C T G G G T T ; Nap2 forward, C C T T C A T A T G G C A G A T C A C A G T T T T T C A G A T G G G G T ; and Nap2 reverse, G A C A A A G C T T A C A C C T T G G G G T T A A T T T C C G C A T C A .

The amplifi ed products were purifi ed (QIAGEN), digested with NdeI and HindIII, and ligated into pHIT51 (containing T7 promoter, D box, His sequence, and multicloning site; provided by H. Tabara, Kyoto University, Kyoto, Japan) digested with the same enzymes. The resulting plasmids were verifi ed by nucleotide sequencing. The expression plasmids for PP2Cγ and the mutants were also constructed by inserting the PP2Cγ se-quence (Murray et al., 1999) into pHIT51. Each plasmid was introduced into BL21-Gold (Stratagene), and expression was induced with 1 mM IPTG at 30°C. The His-tagged proteins were purifi ed using Ni-agarose beads (Sigma-Aldrich) according to the standard protocol by the manufacturer followed by MonoQ chromatography (GE Healthcare) and elution with a linear NaCl gradient (0.1–1 M NaCl in 50 mM Tris-Cl, pH 8.0).

For the phosphatase assay of bulk histones, histones purifi ed from 20 μg HeLa cells were phosphorylated using 1 μg MSK1 (Upstate Biotech-nology) in 12 mM MOPS, pH 7.0, 15 mM MgCl2, 0.2 mM EDTA, 1 mM EGTA, 0.2 mM DTT, 0.1 mM ATP, and 7.4 MBq/ml γ-[32P]ATP for 10 min at 30°C. The unincorporated ATP was removed, and the buffer was substi-tuted with TMD (10 mM Tris-Cl, pH 8.0, 10 mM MgCl2, and 1 mM DTT) using Ultrafree fi lters 0.5 (Millipore). The phosphorylated histones (2 μg in 10 μl) were mixed with a serial dilution of His-tagged proteins (2.5 μl) and incubated for 1 h at 37°C. After stopping the reactions by adding 12.5 μl of 2× SDS gel loading buffer (Sambrook et al., 1989) and boiling, the samples were separated by SDS-PAGE and stained with Coomassie. The radioactivity was detected using an imaging analyzer (BAS2000; Fujifi lm). For the γ-H2AX dephosphorylation assay, the γ-H2AX–containing H2A–H2B fraction was prepared from HeLa cells irradiated (12 Gy) using a 137Cs source at a dose rate of 1.13 Gy/min (Gammacell 40 Exactor; MDS Nordion). The H2A–H2B sample (1 μg in 10 μl) was mixed with 2.5 μl phosphatases in TMD buffer, incubated for 30 min at 37°C, and the level of γ-H2AX was analyzed by immunoblotting using antiphospho-H2AX Ser139 antibody (1:1,000; Upstate Biotechnology).

Immunoprecipitation and AUT gel electrophoresisTo immunoprecipitate GFP-H2A and its binding proteins, 1 ml S100 extract from HeLa cells (control) or cells expressing GFP-H2A was mixed with 50 μl anti-GFP agarose beads (Nacalai Tesque). After incubation for 1.5 h at 4°C with rotation, the beads were collected by centrifugation at 1,600 g for 5 min at 4°C. After washing four times for 10 min at 4°C in PB containing 0.05% Tween 20, 0.2 M NaCl, and protease inhibitor cocktail (Nacalai Tesque), the immunoprecipitates were eluted from the beads by boiling for 10 min in 60 μl of 2× SDS gel loading buffer (Sambrook et al., 1989).

For Fig. 5 (A and B), the FLAG-PP2Cγ, -∆AcDo, and -D496A plas-mids were generated by inserting the corresponding cDNAs into a modi-fi ed version of pcDNA3.1/Hygro (Invitrogen) that contains N-terminal

FLAG and V5 tags. 293T cells (4 × 90-cm2 dishes; 20% confl uent) were transfected with these constructs using a calcium phosphate precipitation method (Sambrook et al., 1989). 3 d later, cells were washed with ice-cold PB, lysed in 2 ml PB containing 0.1% Triton X-100 and protease inhibitor cocktail, incubated for 5 min on ice, and cleared by centrifuga-tion at 1,600 g for 10 min at 4°C. The supernatant was collected and mixed with 100 μl anti-FLAG agarose M2 beads (Sigma-Aldrich). After incubation and washing in the same buffer four times for 10 min at 4°C, the immunoprecipitated material was eluted with 100 μg/ml 3× FLAG peptide in PB (three times at 150 μl). The elution was pooled and either mixed with 2× SDS gel loading buffer for SDS-PAGE or with 20 mg/ml Casamino acids (fi nal concentration of 100 μg/ml; Difco) and 100% tri-chloroacetic acid (fi nal concentration of 20%) for AUT gel electrophoresis. After incubation for 1 h on ice and centrifugation at 20,000 g for 30 min at 4°C, the pellet was washed with acetone chilled at −20°C and dissolved in AUT sample buffer (Pilch et al., 2004). Immunoblotting was performed as described previously (Kimura and Cook, 2001) using the fol-lowing primary antibodies: rabbit antiphospho-H2A/H4 Ser1 (1:1,000; Upstate Biotechnology), mouse antiphospho-H2AX Ser139 (1:1,000; Upstate Biotechnology), and mouse antiphospho-H2B Ser14 (clone 6C9; 1:20 hybridoma supernatant).

To produce antiphospho-H2B Ser14, mice were immunized with a synthetic peptide KSAPAPKKG(phospho-S)K K A V T K A Q K C (Sigma-Genosys) coupled to keyhole limpet hemocyanin (Kimura et al., 1994), and a hybrid-oma clone 6C9 was obtained by ELISA screening using the phosphory-lated and unphosphorylated peptides. As H2B Ser14 is phosphorylated during apoptosis (Cheung et al., 2003), the specifi city was then checked by the specifi c appearance of positive signals in apoptosis-induced (etopo-side treated) HeLa cells by immunoblotting and immunofl uorescence.

siRNA transfection and photobleachingPP2Cγ-specifi c Stealth RNA (Invitrogen; nucleotide number 351-376 or 642-667 of GenBank EMBL/DDBJ accession no. NM_177983) and the control RNA (Invitrogen; number 12935-300) were transfected using Lipo-fectAMINE2000 (Invitrogen). Total cellular proteins were prepared 1–3 d after transfection, separated on an 8% SDS-polyacrylamide gel, and immuno-blotted (Kimura and Cook, 2001) with mouse monoclonal antibody di-rected against PP2Cγ (1:10,000; Murray et al., 1999) or α-tubulin (1:1,000; Oncogene Research Products) as a control. Cells grown on cov-erslips were transfected with Stealth RNA and fi xed for immunofl uores-cence using the mouse anti-PP2Cγ (1:30,000) and Cy3-conjugated anti–mouse IgG (1:500; Jackson ImmunoResearch Laboratories).

For photobleaching studies, HeLa cells expressing GFP-H2A, H2B-GFP, or H1c-GFP (Misteli et al., 2000) grown on glass-bottom dishes (Mat-Tek) were transfected with Stealth RNA. 3 d later, the dish was set on an inverted microscope (LSM510 META; Carl Zeiss MicroImaging, Inc.) in an air chamber at 37°C, and the mobility was analyzed by photobleaching using the inverted microscope with a plan-Neofl uar 40× NA 1.3 objective. For H2A and H2B, fi ve confocal images were collected (512 × 512 pix-els, zoom 3, maximum scan speed, pinhole 3.7 airy unit, LP505 emission fi lter, and 0.3% transmission of 458-nm Ar laser with 75% output power), one half of a nucleus was bleached using 100% transmission of 458 and 488 nm (eight iterations), and images were collected using the original set-ting every 5 min. For H1c, fi ve images were collected (256 × 256 pixels, zoom 8, and scan speed 12), a 2-μm spot was bleached using 100% transmission of 458 and 488 nm (eight iterations), and images were col-lected every 5 s (the graph in Fig. 7 E shows the points of every 10 s for ease of comparison). The fl uorescence intensity of the bleached area was measured using MetaMorph software (Molecular Devices). After subtract-ing the background, the intensity was normalized to the initial intensity before bleaching.

DT40 cellsPP2Cγ-defi cient DT40 cells were established using standard methods (Fig. S3; Fukagawa et al., 2004) and grown at 37°C. To measure the cell den-sity, cells were mixed with trypan blue solution (Invitrogen), and the number of live cells excluding the dye was counted. To determine the sensitivity to caffeine and irradiation, serially diluted cells were plated in methylcellu-lose plates with or without 1 mM caffeine (Sigma-Aldrich) and irradiated using a Gammacell 40 Exactor (Nordion). Colonies were counted 10–12 d after plating. For immunoblotting (Fig. 8 D), 4 × 105 cells/ml were irradi-ated, and calyculin A (Sigma-Aldrich) was immediately added (fi nal con-centration of 10 ng/ml). A 1-ml aliquot was taken at each time point, and cells were collected (600 g for 2 min) and lysed in 100 μl of 2× SDS gel loading buffer.

on Decem


Dow

nloaded from




Online supplemental materialFig. S1 shows that ATP is not required for GFP-H2A incorporation into chromatin in permeabilized cells assisted by PP2Cγ or Nap1. Fig. S2 shows that PP2Cγ has weak de novo nucleosome assembly activity. Fig. S3 shows evidence for the generation of PP2Cγ knockout DT40 cells. Online supplemental material is available at http://www.jcb.org/cgi/content/full/jcb.200608001/DC1.

We thank D.T. Brown, Y. Ishimi, T. Kanda, P.Y. Perche, H. Tabara, C. Vourc’h, and G. Wahl for materials, and Y. Agata, T. Ikura, H. Kurumizaka, T. Misteli, and S. Tashiro for valuable discussion and comments on the manuscript.

This work was prepared, in part, at the Radiation Biology Center and the Radioisotope Research Center (Kyoto University). This work was supported by Grants-in-aid from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan and the Special Coordination Funds for Promoting Science and Technology from the MEXT of Japan. E. Allemand and A.R. Krainer acknowledge support from the National Institutes of Health grant GM42699.

Submitted: 1 August 2006Accepted: 25 September 2006

ReferencesAbraham, R.T. 2004. PI 3-kinase related kinases: ‘big’ players in stress-induced

signaling pathways. DNA Repair (Amst.). 3:883–887.

Batsche, E., M. Yaniv, and C. Muchardt. 2006. The human SWI/SNF subunit Brm is a regulator of alternative splicing. Nat. Struct. Mol. Biol. 13:22–29.

Barman, H.K., Y. Takami, T. Ono, H. Nishijima, F. Sanematsu, K. Shibahara, and T. Nakayama. 2006. Histone acetyltransferase 1 is dispensable for repli-cation-coupled chromatin assembly but contributes to recover DNA dam-ages created following replication blockage in vertebrate cells. Biochem. Biophys. Res. Commun. 345:1547–1557.

Belotserkovskaya, R., and D. Reinberg. 2004. Facts about FACT and transcript elongation through chromatin. Curr. Opin. Genet. Dev. 14:139–146.

Benson, L.J., Y. Gu, T. Yakovleva, K. Tong, C. Barrows, C.L. Strack, R.G. Cook, C.A. Mizzen, and A.T. Annunziato. 2006. Modifi cations of H3 and H4 during chromatin replication, nucleosome assembly, and histone exchange. J. Biol. Chem. 281:9287–9296.

Bruno, M., A. Flaus, C. Stockdale, C. Rencurel, H. Ferreira, and T. Owen-Hughes. 2003. Histone H2A/H2B dimer exchange by ATP-dependent chromatin remodeling activities. Mol. Cell. 12:1599–1606.

Chang, L., S.S. Loranger, C. Mizzen, S.G. Ernst, C.D. Allis, and A.T. Annunziato. 1997. Histones in transit: cytosolic histone complexes and diacetylation of H4 during nucleosome assembly in human cells. Biochemistry. 36:469–480.

Cheung, W.L., K. Ajiro, K. Samejima, M. Kloc, P. Cheung, C.A. Mizzen, A. Beeser, L.D. Etkin, J. Chernoff, W.C. Earnshaw, and C.D. Allis. 2003. Apoptotic phosphorylation of histone H2B is mediated by mammalian sterile twenty kinase. Cell. 113:507–517.

Chowdhury, D., M.C. Keogh, H. Ishii, C.L. Peterson, S. Buratowski, and J. Lieberman. 2005. γ-H2AX dephosphorylation by protein phosphatase 2A facilitates DNA double-strand break repair. Mol. Cell. 20:801–809.

Dignam, J.D., R.M. Lebovits, and R.G. Roeder. 1983. Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mam-malian nuclei. Nucleic Acids Res. 11:1475–1489.

Fernandez-Capetillo, O., C.D. Allis, and A. Nussenzweig. 2004. Phosphorylation of histone H2B at DNA double-strand breaks. J. Exp. Med. 199:1671–1677.

Flaus, A., and T. Owen-Hughes. 2004. Mechanisms for ATP-dependent chro-matin remodelling: farewell to the tuna-can octamer? Curr. Opin. Genet. Dev. 14:165–173.

Fukagawa, T., M. Nogami, M. Yoshikawa, M. Ikeno, T. Okazaki, Y. Takami, T. Nakayama, and M. Oshimura. 2004. Dicer is essential for formation of the heterochromatin structure in vertebrate cells. Nat. Cell Biol. 6:784–791.

Hake, S.B., B.A. Garcia, E.M. Duncan, M. Kauer, G. Dellaire, J. Shabanowitz, D.P. Bazett-Jones, C.D. Allis, and D.F. Hunt. 2006. Expression patterns and post-translational modifi cations associated with mammalian histone H3 variants. J. Biol. Chem. 281:559–568.

Henikoff, S., and K. Ahmad. 2005. Assembly of variant histones into chromatin. Annu. Rev. Cell Dev. Biol. 21:133–153.

Ishimi, Y., J. Hirosumi, W. Sato, K. Sugasawa, S. Yokota, F. Hanaoka, and M. Yamada. 1984. Purifi cation and initial characterization of a protein which facilitates assembly of nucleosome-like structure from mammalian cells. Eur. J. Biochem. 142:431–439.

Jackson, V. 1990. In vivo studies on the dynamics of histone-DNA interaction: evidence for nucleosome dissolution during replication and transcrip-tion and a low level of dissolution independent of both. Biochemistry. 29:719–731.

Jackson, D.A., and P.R. Cook. 1985. A general method for preparing chromatin containing intact DNA. EMBO J. 4:913–918.

Jackson, S., W. Brooks, and V. Jackson. 1994. Dynamics of the interactions of his-tones H2A,H2B and H3,H4 with torsionally stressed DNA. Biochemistry. 33:5392–5403.

Kanda, T., K.F. Sullivan, and G.M. Wahl. 1998. Histone-GFP fusion protein enables sensitive analysis of chromosome dynamics in living mammalian cells. Curr. Biol. 8:377–385.

Kaufmann, W.K., T.P. Heffernan, L.M. Beaulieu, S. Doherty, A.R. Frank, Y. Zhou, M.F. Bryant, T. Zhou, D.D. Luche, N. Nikolaishvili-Feinberg, et al. 2003. Caffeine and human DNA metabolism: the magic and the mystery. Mutat. Res. 532:85–102.

Keogh, M.C., J.A. Kim, M. Downey, J. Fillingham, D. Chowdhury, J.C. Harrison, M. Onishi, N. Datta, S. Galicia, A. Emili, et al. 2005. A phosphatase com-plex that dephosphorylates γH2AX regulates DNA damage checkpoint recovery. Nature. 439:497–501.

Kimura, H. 2005. Histone dynamics in living cells revealed by photobleaching. DNA Repair (Amst.). 4:939–950.

Kimura, H., and P.R. Cook. 2001. Kinetics of core histones in living cells: little exchange of H3 and H4 and some rapid exchange of H2B. J. Cell Biol. 153:1341–1353.

Kimura, K., N. Nozaki, M. Saijo, A. Kikuchi, M. Ui, and T. Enomoto. 1994. Identifi cation of the nature of modifi cation that causes the shift of DNA topoisomerase II beta to apparent higher molecular weight forms in the M phase. J. Biol. Chem. 269:24523–24526.

Kireeva, M.L., W. Walter, V. Tchernajenko, V. Bondarenko, M. Kashlev, and V.M. Studitsky. 2002. Nucleosome remodeling induced by RNA polymerase II: loss of the H2A/H2B dimer during transcription. Mol. Cell. 9:541–552.

Kobor, M.S., S. Venkatasubrahmanyam, M.D. Meneghini, J.W. Gin, J.L. Jennings, A.J. Link, H.D. Madhani, and J. Rine. 2004. A protein complex containing the conserved Swi2/Snf2-related ATPase Swr1p deposits his-tone variant H2A.Z into euchromatin. PLoS Biol. 2:E131.

Krogan, N.J., K. Baetz, M.C. Keogh, N. Datta, C. Sawa, T.C. Kwok, N.J. Thompson, M.G. Davey, J. Pootoolal, T.R. Hughes, et al. 2004. Regulation of chromosome stability by the histone H2A variant Htz1, the Swr1 chromatin remodeling complex, and the histone acetyltransferase NuA4. Proc. Natl. Acad. Sci. USA. 101:13513–13518.

Kusch, T., L. Florens, W.H. Macdonald, S.K. Swanson, R.L. Glaser, J.R. Yates III, S.M. Abmayr, M.P. Washburn, and J.L. Workman. 2004. Acetylation by Tip60 is required for selective histone variant exchange at DNA lesions. Science. 306:2084–2087.

Leroy, C., S.E. Lee, M.B. Vaze, F. Ochsenbein, R. Guerois, J.E. Haber, and M.C. Marsolier-Kergoat. 2003. PP2C phosphatases Ptc2 and Ptc3 are required for DNA checkpoint inactivation after a double-strand break. Mol. Cell. 11:827–835.

Louters, L., and R. Chalkley. 1985. Exchange of histones H1, H2A, and H2B in vivo. Biochemistry. 24:3080–3085.

Loyola, A., and G. Almouzni. 2004. Histone chaperones, a supporting role in the limelight. Biochim. Biophys. Acta. 1677:3–11.

Lu, X., B. Nannenga, and L.A. Donehower. 2005. PPM1D dephosphory-lates Chk1 and p53 and abrogates cell cycle checkpoints. Genes Dev. 19:1162–1174.

Luger, K., A.W. Mader, R.K. Richmond, D.F. Sargent, and T.J. Richmond. 1997. Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature. 389:251–260.

Ma, X.J., J. Wu, B.A. Altheim, M.C. Schultz, and M. Grunstein. 1998. Deposition-related sites K5/K12 in histone H4 are not required for nucleosome depo-sition in yeast. Proc. Natl. Acad. Sci. USA. 95:6693–6698.

Maison, C., D. Bailly, A.H.F.M. Peters, J.P. Quivy, D. Roche, A. Taddei, M. Lachner, T. Jenuwein, and G. Almouzni. 2002. Higher-order structure in pericentric heterochromatin involves a distinct pattern of histone modifi -cation and an RNA component. Nat. Genet. 30:329–334.

Manser, T., T. Thacher, and M. Rechsteiner. 1980. Arginine-rich histones do not exchange between human and mouse chromosomes in hybrid-cells. Cell. 19:993–1003.

Meshorer, E., D. Yellajoshula, E. George, P.J. Scambler, D.T. Brown, and T. Misteli. 2006. Hyperdynamic plasticity of chromatin proteins in pluripo-tent embryonic stem cells. Dev. Cell. 10:105–116.

Misteli, T., and D.L. Spector. 1996. Serine/threonine phosphatase 1 modulates the subnuclear distribution of pre-mRNA splicing factors. Mol. Biol. Cell. 7:1559–1572.

Misteli, T., A. Gunjan, R. Hock, M. Bustin, and D.T. Brown. 2000. Dynamic binding of histone H1 to chromatin in living cells. Nature. 408:877–881.

on Decem


Dow

nloaded from




Mizuguchi, G., X. Shen, J. Landry, W.H. Wu, S. Sen, and C. Wu. 2004. ATP-driven exchange of histone H2AZ variant catalyzed by SWR1 chromatin remodeling complex. Science. 303:343–348.

Murray, M.V., R. Kobayashi, and A.R. Krainer. 1999. The type 2C Ser/Thr phos-phatase PP2Cγ is a pre-mRNA splicing factor. Genes Dev. 13:87–97.

Nazarov, I.B., A.N. Smirnova, R.I. Krutilina, M.P. Svetlova, L.V. Solovjeva, A.A. Nikiforov, S.L. Oei, I.A. Zalenskaya, P.M. Yau, E.M. Bradbury, and N.V. Tomilin. 2003. Dephosphorylation of histone gamma-H2AX during re-pair of DNA double-strand breaks in mammalian cells and its inhibition by calyculin A. Radiat. Res. 160:309–317.

Ohkuni, K., K. Shirahige, and A. Kikuchi. 2003. Genome-wide expression analysis of NAP1 in Saccharomyces cerevisiae. Biochem. Biophys. Res. Commun. 306:5–9.

Park, Y.J., J.V. Chodaparambil, Y. Bao, S.J. McBryant, and K. Luger. 2005. Nucleosome assembly protein 1 exchanges histone H2A-H2B dimers and assists nucleosome sliding. J. Biol. Chem. 280:1817–1825.

Perche, P.Y., C. Vourc’h, L. Konecny, C. Souchier, M. Robert-Nicoud, S. Dimitrov, and S. Khochbin. 2000. Higher concentrations of histone mac-roH2A in the Barr body are correlated with higher nucleosome density. Curr. Biol. 10:1531–1534.

Pilch, D.R., C. Redon, O.A. Sdelnikova, and W.M. Bonner. 2004. Two- dimensional gel analysis of histones and other H2AX-related methods. Methods Enzymol. 375:76–88.

Pombo, A., D.A. Jackson, M. Hollinshead, Z. Wang, R.G. Roeder, and P.R. Cook. 1999. Regional specialization in human nuclei: visualization of discrete sites of transcription by RNA polymerase III. EMBO J. 18:2241–2253.

Rodriguez, P., D. Munroe, D. Prawitt, L.L. Chu, E. Bric, J. Kim, L.H. Reid, C. Davies, H. Nakagama, R. Loebbert, et al. 1997. Functional characteriza-tion of human nucleosome assembly protein-2 (NAP1L4) suggests a role as a histone chaperone. Genomics. 44:253–265.

Rogakou, E.P., C. Boon, C. Redon, and W.M. Bonner. 1999. Megabase chroma-tin domains involved in DNA double-strand breaks in vivo. J. Cell Biol. 146:905–916.

Maniatis, T., E.F. Fritsch, and J. Sambrook. 1989. Molecular Cloning: A Laboratory Manual. 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 545 pp.

Simon, R.H., and G. Felsenfeld. 1979. A new procedure for purifying histone pairs H2A + H2B and H3 + H4 from chromatin using hydroxylapatite. Nucleic Acids Res. 6:689–696.

Tagami, H., D. Ray-Gallet, G. Almouzni, and Y. Nakatani. 2004. Histone H3.1 and H3.3 complexes mediate nucleosome assembly pathways dependent or independent of DNA synthesis. Cell. 116:51–61.

Thiriet, C., and J.J. Hayes. 2005. Replication-independent core histone dynamics at transcriptionally active loci in vivo. Genes Dev. 19:677–682.

Travis, S.M., and M.J. Welsh. 1997. PP2Cγ: a human protein phosphatase with a unique acidic domain. FEBS Lett. 412:415–419.

Turner, B.M. 2002. Cellular memory and the histone code. Cell. 111:285–291.

on Decem


Dow

nloaded from



A novel histone exchange factor, protein γ, mediates the exchange …repository.cshl.edu/22839/1/A...

Documents

Transcript of A novel histone exchange factor, protein γ, mediates the exchange …repository.cshl.edu/22839/1/A...